An important finding of this study was that the presence of bone and/or CNS metastases at recurrence was predictive of a poor prognosis in EGFR + NSCLC patient with PR Furthermore, the pattern of recurrence was not identified as exerting any significant influence on the prognosis.
In NSCLC patients with PR the presence or absence of EGFR mutation, number and sites of recurrence (intra/extra thoracic), and presence/absence of CNS metastasis have been well-established as prognostic factors (15–16), although there are conflicting opinions regarding the influence of bone metastases at recurrence on the prognosis (7–8). While the presence of bone metastasis at recurrence is reported as a poor prognostic factor in terms of both the PFS and OS in patients with advanced EGFR-positive NSCLC (17), there are no reports until date on the prognostic impact of bone metastasis in EGFR + NSCLC patients with PR. According to a previous report, the reason for the poor prognosis associated with bone metastasis in patients with advanced EGFR-positive NSCLC is the presence, at a high frequency (72.7, of multiple (3 or more) distant metastases in these patients (17). In the present study, among the 33 patients with PR, including those with bone metastasis, 12 (36.4%) had bone metastases only and 21 (63.6%) had other distant metastases in addition, showing a tendency for the presence of other distant metastases at a high frequency in patients with bone metastasis; this was consistent with the aforementioned report (17), although the number of coexistent metastases differed between the two studies. The number of patients who survived for more than 10 years after the diagnosis of recurrence was 20 (12.3%), including 8 patients with pathological stage I disease, 4 with pathological stage II disease, and 8 with pathological stage III disease. Only one of these patients had bone metastasis, two had CNS metastasis (in one site of the CNS in 1 patient, and in 2 sites in the other patient), and about a half of these patients (11 patients) had oligo recurrence.
Numerous studies have reported the relatively high risk of bone and CNS metastases in NSCLC patients with PR (18–19). According to one previous report, women are less likely to develop bone metastasis (18). Also, associations of the preoperative carcinoembryonic antigen levels (> 5 ng/mL) and positive pathologic lymph nodes with the development of bone metastasis, and of poorly differentiated tumor and positive pathologic lymph nodes with the development of CNS metastasis have been reported (20). Bone metastasis and male gender are thought to be associated regardless of the presence/absence of EGFR gene mutations, but further investigation of the tumor markers and pathological findings and the risk of bone and CNS metastases in postoperative EGFR-positive NSCLC patients is needed. In the ADAURA trial, a subgroup analysis showed a trend toward worse OS in male patients as compared with female patients (HR 0.62 [95% CI, 0.33–1.13] for male patients and HR 0.41 [95% CI, 0.25–0.66] for female patients) (5). In this study also, male gender was associated with the presence of bone and CNS metastases, and new treatment strategies are needed for patients with bone and CNS metastases at first recurrence.
Among the NSCLC patients with PR the PRS was higher in the oligo recurrence group than in the non oligo recurrence group, and local therapy in the oligo group improved the PRS further (9). However, local therapy has been reported as exerting no influence on the PRS in EGFR + NSCLC patients with PR as TKIs are very effective in this patient group (21). In the present study, 52 patients (78%) of the oligo recurrence group received local therapy, such as chemoradiotherapy, radiotherapy or resection, but neither the treatment strategy used nor the pattern of recurrence had any influence on the prognosis, as previously reported (21). Because the number of patients taking osimertinib developing recurrence could be expected to increase, further studies are needed to develop effective treatments for postoperative NSCLC patients with bone/CNS metastasis at first recurrence. A randomized controlled phase III trial comparing systemic versus local therapy for postoperative oligo recurrence in NSCLC patients is currently being planned (JCOG2108).
The reported PRS in EGFR + NSCLC patients with PR treated with TKIs is 4.08 years (8), and shorter than that determined in this study. This could be attributable to the larger number of patients treated with osimertinib in this study. In both the osimertinib and placebo arms, the 5-year survival rates were higher in the ADAURA trial than in this study. This result could be explained by the fact that 79 (38.5%) of the 205 patients in the placebo group who showed disease relapse in the ADAURA trial received osimertinib as the primary treatment after recurrence (5), as compared to the 23 (14.2%) patients in the present study, and that we only enrolled patients with PR in this study. In the 23 patients in this study who received osimertinib as first-line therapy after recurrence, the PRS was still NA at the time of writing (95% CI NA-NA) and the 5-year PRS was 100% (95% CI NA-NA). This result lends support to the rationale of using osimertinib as the first-line therapy after postoperative recurrence. Although it was demonstrated in the ADAURA trial that osimertinib treatment was associated with a reduced incidence of bone and CNS metastases (22), some patients still hesitate to take adjuvant osimertinib treatment owing to the reported adverse effects. In the ADAURA trial, 36 patients in the osimertinib group developed adverse events, and in 30 of these, the treatment was discontinued at the patient's request (4). We consider that the findings of our present study are useful for patients considering resuming osimertinib.
This study had several limitations. Firstly, it was a retrospective study conducted at a single institution, and the possibility of selection bias cannot be ruled out. Secondly, histological confirmation was not obtained at recurrence in all patients. Thirdly, we did not compare the prognostic factors with those in NSCLC patients with wild-type EGFR. Fourthly, osimertinib was not used in all patients. The FLAURA trial showed that osimertinib prevented progression of CNS metastases significantly more effectively than first- or second-generation TKIs in advanced EGFR-positive NSCLC patients (23); therefore, the use of osimertinib may significantly improve the prognosis in patients with CNS metastasis. Also, in the ADAURA trial, osimertinib treatment yielded better CNS DFS in patients with stage II-IIIA disease (HR 0.24 [95% CI 0.14–0.42]) and a reduced incidence of bone metastasis (13 vs. 32 patients) (22). Because patients treated with osimertinib as adjuvant therapy is expected to increase, the poor prognosis population in this study could be expected to decrease. But we expect that the results of our present study will be helpful for countries in which osimertinib is still not approved as adjuvant therapy for NSCLC, and for patients who refuse to take osimertinib.